Even with the history of debate over the merits of homework, there are significant gaps in the research record regarding its benefit to students. The focus of this study is on the association between time spent on homework and academic performance in science and math by assessing survey and transcript data from two nationally representative samples of high school students collected in 1990 and 2002. Using multiple linear regressions and controlling for students' background, motivation, and prior achievement, we investigated how much variance in science and math course grades and achievement test scores could be explained by time spent on homework in those classes. The results indicate that there is no consistent significant relationship between time spent on homework and grades, but a consistently positive significant relationship between homework and performance on standardized exams. (Contains 7 tables and 3 footnotes.)